Abstarct: In statistics, for estimating the density function, nonparametric methods can be applied. On the other hand, it is clear that the independency feature is less accepted in actual conditions, and dependence can be observed between most phenomena in nature. In situations where the risk analysis plays a role, improved estimators often improve the estimation process by using prior information on some or all of the model's parameters understudy from risk sense. Therefore, in this thesis, using non-sample information, we study the shrinkage estimators for the density function, assuming specific associated dependence structures. Also, considering the α-mixing dependency structure, we introduce the improved estimators related to the derivative of the density function and the hazard rate function, respectively, and examine their asymptotic properties and the performance of these proposed improved estimators relative to the corresponding kernel estimator.